发明名称 Enabling cluster scaling
摘要 Embodiments of the present invention relate to methods and apparatuses for enabling cluster scaling. Specifically, there is disclosed a method of recording a context of configuration for an initial node of a cluster, comprising: retrieving at least one file to be used for configuring the initial node; and recording a context while configuring the initial node using the at least one file, the recorded context being information on the configuration of the initial node and to be used for enabling addition of a new node into the cluster. There is also disclosed a method of enabling addition of a new node into a cluster. Corresponding apparatuses are also disclosed. According to embodiments of the present invention, the efficiency of improving cluster scaling may be effectively improved.
申请公布号 US9288109(B2) 申请公布日期 2016.03.15
申请号 US201213612445 申请日期 2012.09.12
申请人 INTERNATIONAL BUSINESS MACHINES CORPORATION 发明人 Jian Li Rong;Qiu Jie;Yang Jie Y.;Yu Tao;Zhong Xiao
分类号 G06F15/177;H04L12/24;H04L29/08 主分类号 G06F15/177
代理机构 Scully, Scott, Murphy & Presser, P.C. 代理人 Scully, Scott, Murphy & Presser, P.C. ;Davis, Esq. Jennifer
主权项 1. A method of recording a context of configuration for an initial node of a cluster, comprising: retrieving at least one file to be used for configuring the initial node; recording, at the initial node, a context of the initial node, the recording being performed while configuring the initial node using the at least one file, the recorded context being information on the configuration of the initial node and to be used for enabling addition of a new node into the cluster, the context indicating at least one file to be retrieved first from a file repository, at least one file to be retrieved from the initial node, a sequence of the files executed in the new node being added to the cluster, and dependencies indicating whether a retrieved file is to be run dependent upon execution of a preceding file; and configuring the new node by concurrently retrieving files without a dependency, said configuring including running a retrieved file according to the sequence while concurrently retrieving a next file without a dependency from the file repository; wherein the new node determines whether to record its own context; and wherein the new node determines whether to record its own context based upon how many other sources of context are currently available in the cluster.
地址 Armonk NY US
您可能感兴趣的专利